From d1e0a4607d3914cda9e425d2ac5f240ca17b757d Mon Sep 17 00:00:00 2001
From: Ross Wightman <rwightman@gmail.com>
Date: Mon, 21 Nov 2022 17:42:19 -0800
Subject: [PATCH] Switch TFDS dataset to use INTEGER_ACCURATE jpeg decode by
 default

---
 timm/data/readers/reader_tfds.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/timm/data/readers/reader_tfds.py b/timm/data/readers/reader_tfds.py
index a327df7b..25aab471 100644
--- a/timm/data/readers/reader_tfds.py
+++ b/timm/data/readers/reader_tfds.py
@@ -43,6 +43,15 @@ SHUFFLE_SIZE = int(os.environ.get('TFDS_SHUFFLE_SIZE', 8192))  # samples to shuf
 PREFETCH_SIZE = int(os.environ.get('TFDS_PREFETCH_SIZE', 2048))  # samples to prefetch
 
 
+@tfds.decode.make_decoder()
+def decode_example(serialized_image, feature, dct_method='INTEGER_ACCURATE'):
+    return tf.image.decode_jpeg(
+        serialized_image,
+        channels=3,
+        dct_method=dct_method,
+    )
+
+
 def even_split_indices(split, n, num_samples):
     partitions = [round(i * num_samples / n) for i in range(n + 1)]
     return [f"{split}[{partitions[i]}:{partitions[i + 1]}]" for i in range(n)]
@@ -242,6 +251,7 @@ class ReaderTfds(Reader):
         ds = self.builder.as_dataset(
             split=self.subsplit or self.split,
             shuffle_files=self.is_training,
+            decoders=dict(image=decode_example()),
             read_config=read_config,
         )
         # avoid overloading threading w/ combo of TF ds threads + PyTorch workers